Required when terminal_* tools are available. Explains foreground execution, outer collect_result handles, promoted job IDs and their retrieval/cancellation, deadlines, output retention, platform shell selection, and structured editing when enabled.
The skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
The required workflow does not read an outsider-authored queue, but the skill routes browser and web-search tasks to arbitrary web pages, which the agent must actively fetch and read (low).
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.